Code for MPC Algorithm

This repository contains a reference implementation of the algorithms for the paper:

Hongchao Qin, Rong-Hua Li, Guoren Wang, Lu Qin, Yurong Cheng, Ye Yuan. Mining Periodic Cliques in Temporal Networks. ICDE 2019: 1130-1141

Environment Setup

Codes run on Python 2.7 or later. PyPy compiler is recommended because it can make the computations quicker without change the codes.

You may use Git to clone the repository from GitHub and run it manually like this:

git clone https://github.com/VeryLargeGraph/MPC.git
cd MPC
pip install click
python run.py

Dateset description

We focus on mining the temporal network so each edge is associated with a timestamp. Temporal edges are stored at the raw data in which each line is one temporal edge.

from_id	\t	to_id	\t	timestamps

Note that, the function readGraph in mpc.py can make each snapshot G_i (see Fig 1.c in the paper) to be a simple graph.

Running example

You can type in dataset name, parameters theta, k and method name to control the program:

Dataset name(str): chess_year
theta(int): 3
k(int): 3
Type one number to chose the algorithm: [1]MPCKC; [2]MPCWC; [3]MPCSC. (int): 1
loading...
number of nodes:7301
number of edges:55899
number of temporal edges:62385
kCore time:0:00:00.013000
kCore #nodes:5472
#new nodes:8416
#new edges:681
#mpc:3
All time:0:00:00.199000
New nodes stored in file "NEWNODES.json" : {new_node_id: [raw_node_id, [starttime, interval]]}
if theta = 4, one node [1, [2003, 1]] means node id 1 in raw data is periodic at 2003, 2004, 2005, 2006
if theta = 3, one node [2, [2003, 2]] means node id 2 in raw data is periodic at 2003, 2005, 2007
MPCliques (New nodes):
[set([302, 263, 742, 762]), set([1840, 1848, 3448, 3520]), set([3872, 4151, 4132, 7479])]